In Simple Words
Imagine buying rain insurance that relies on a weather gauge located twenty miles away at a city airport. A localized drought destroys crops on a farm, but because normal rain fell at the distant airport station, the insurance company refuses to pay out while keeping the farmer's premium.
Villages considered for the pilot
- Mainly drought risk
- 84 passed the hazard check
- Mainly frost or flood risk
- No suitable index available
Observed village selection, not a treatment effect. Drought villages faced a further rainfall-record check before inclusion.
See source →Research question
Weather index insurance fails to deliver effective financial protection when rainfall measurements collected at regional weather stations do not correlate accurately with actual crop losses on individual smallholder plots.
Participants
120 Kebeles initially selected across 4 zones in Amhara, Ethiopia; reduced to 49 study Kebeles covered by 17 rainfall stations.
Study Design
Meteorological station availability and spatial coverage analysis combined with theoretical risk models of index contract design.
Results
36 out of 120 Kebeles (30%) were excluded because dominant risks were non-drought perils (frost and flooding) lacking indices, and another 35 Kebeles were dropped due to incomplete historical station rainfall records required for reinsurance pricing.
Takeaway
Establish dense weather monitoring networks or satellite remote-sensing indices to minimize spatial and peril mismatch before offering index insurance products to farmers.